Nanoscale

View Article Online

FEATURE ARTICLE View Journal | View Issue

Growing metal nanoparticles in superfluid helium

Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.

Cite this: Nanoscale, 2013, 5, 11545

Shengfu Yang,*a Andrew M. Ellis,*a Daniel Spence,a Cheng Feng,a
Adrian Boatwright,a Elspeth Latimera and Chris Binnsb

Received 1st August 2013
Accepted 27th September 2013
DOI: 10.1039/c3nr04003h
www.rsc.org/nanoscale
Helium droplets provide a cold and confined environment where atomic and/or molecular dopants can
aggregate into clusters and nanoparticles. In particular, the sequential addition of different materials to
helium droplets can lead to the formation of a wide range of nanoparticles, including core–shell
nanoparticles, which can then be deposited onto a surface. Here we briefly discuss the fundamental
properties of helium droplets and then address their implications for the formation of clusters and
nanoparticles. Several key experiments on atomic and molecular clusters will be highlighted and new
results obtained for nanoparticles formed in this way will be presented. Finally, the versatility, the
limitations and new possibilities provided by superfluid helium droplets in nanoscience and
nanotechnology will be addressed.

1. Introduction nanoparticles. Further manipulation and assembly of the iso-

When compared with bulk materials, nanoparticles have
distinct and fascinating properties that can depend strongly on
their size, shape, internal structure and chemical composition.
This means that nanoparticles can in principle be ‘tuned’ for
specic applications and many exciting possibilities have
already been identied in medicine,1,2 electronic and opto-
electronic devices,3,4 catalysis,5–7 and even consumer products
such as clothing and sunscreen creams.8,9 As a result, nano-
particles have now become the focus of a huge scientic and
technological research effort and major challenges include the
high degree of control over the growth of nanoparticles in order
to achieve the desired properties and performance, as well as
achieving a good understanding of fundamental materials
science at the nanoscale.
Nanostructures can be fabricated through bottom-up and
top-down approaches. With top-down methods large objects are
modied into nanostructures, for example, by chemical
lithography, plasma methods, beam techniques (electron beam,
ion beam, and lasers) and mechanical processing. Many top-
down approaches have been developed to produce well-dened
structures for industrial applications. However, top-down
methods generally have severe limitations for producing struc-
tures in the size range <10 nm and it is a challenge to create
elaborate 3-D features. On the other hand the bottom-up
approach can produce very small structures by self-assembly of
atoms and molecules, which are bound together by intermo-
lecular forces into clusters and even larger objects, including
lated nanoparticles into nano-devices can also be achieved, for
example, by DNA engineering of nanoparticle super-lattices.10
Such approaches allow production of well-dened nanoparticle
assemblies. Nevertheless, the major challenge for bottom-up
techniques is the precise control of the growth of the nano-
particles. In particular there are severe limitations on the types
of materials that can be incorporated into nanoparticles,
whether alone or in combination with others (e.g. in core–shell
particles), when using conventional synthetic methods such as
wet chemical techniques11,12 or gas aggregation.13,14
In this report we address an emerging bottom-up technique
for the synthesis of nanoparticles, the use of superuid helium
droplets as nano-reactors. This approach offers some degree of
control on the size of the particles but is particularly exciting
because of the wide range of materials that can be incorporated
into these nanoparticles when grown in this environment.
Unlike other review articles and book chapters,15–20 where the
emphasis has been on the broad utility of helium droplets in
physical chemistry, here we focus specically on the synthesis
of metal clusters and nanoparticles using superuid helium.
The review begins with a description of some of the basic
properties of helium droplets that are particularly important in
a cluster and nanoparticle context, and then proceeds to iden-
tify the capabilities and limitations through a discussion of
several recent studies. Finally, we turn to the future and ponder
some remaining challenges that may lead to exciting develop-
ments in nanoscience and nanotechnology using superuid
helium droplets.

a
Department of Chemistry, University of Leicester, Leicester LE1 7RH, UK. E-mail: 2. Fundamentals of helium droplets
sfy1@le.ac.uk; andrew.ellis@le.ac.uk; Fax: +44 (0)116 252 3789; Tel: +44 (0)116
4
252 2127 He is the most abundant isotope of helium and is by far the
b
Department of Physics and Astronomy, University of Leicester, Leicester LE1 7RH, UK most commonly used isotope in helium droplet experiments.

This journal is ª The Royal Society of Chemistry 2013 Nanoscale, 2013, 5, 11545–11553 | 11545

Nanoscale
Unlike other elements in our universe, helium does not possess
a triple point; instead, it has a so-called l-line which separates
the normal liquid phase (He-I), the superuid phase (He-II) and
the solid phase in the phase diagram of 4He. The superuid
phase occurs below 2.18 K,21 while above this temperature
helium behaves as a normal liquid. Superuid helium has a
vanishingly small viscosity and an ultrahigh thermal conduc-
tivity. In addition, superuid helium shows bizarre behaviour
such as the fountain effect, lm ow and creep, and quantized
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
vortices.22,23 The ability to form the superuid phase derives
from the bosonic nature of 4He atoms and at a sufficiently low
temperature the system can form a single macroscopic wave-
function in which individual atom identities are lost. The
superuidity is the consequence of a partial Bose–Einstein
condensation of the liquid helium.
Helium droplets are large helium clusters typically
composed of 103 to 106 helium atoms,18 although even larger
droplets are accessible.24 The most commonly used way to
produce nanoscale helium droplets is via a supersonic expan-
sion of pre-cooled gaseous helium through a pinhole nozzle,
with a 2–5 mm diameter aperture.15,25 The expanding helium gas
in the jet is cooled further by the adiabatic change in state along
isentropes in the phase diagram to the nal temperature and
pressure and begins to form helium droplets.18 At present most
existing helium droplet sources use a continuous expansion of
helium gas15,17 into high vacuum but expansion from the liquid
phase is also possible.18,26 Generally a helium droplet source
consists of a cryostat for initial cooling of the helium, a pinhole
aperture, a skimmer to create a collimated helium droplet
beam, differential pumping to avoid excess collisions that can
destroy the helium droplets, and diagnostic apparatus such as a
mass spectrometer.
Unlike bulk liquid helium, whose temperature can be
continuously adjusted up or down using an appropriate cryo-
stat, the temperature of superuid helium droplets is xed.
Aer their initial formation helium droplets undergo rapid
evaporative loss of helium atoms until a sufficiently low
temperature is reached such that further evaporation is
impossible. Each helium atom can take away a minimum
energy equal to the binding energy of a helium atom to bulk
helium, which is close to 5 cm
1.18 The resulting steady state
temperature is known to be close to 0.38 K, as conrmed by
determining the rotational temperatures of molecules located
inside helium droplets using spectroscopy.27 3He droplets are
even more weakly bound than 4He droplets and therefore have a
lower steady-state temperature of 0.15 K. However, since bulk
3
He only becomes superuid at temperatures below 0.003 K (ref.
28) we would not expect 3He droplets to be superuid. On the
other hand, the superuid phase transition for bulk 4He occurs
below 2.18 K, so it would not be unreasonable to expect
superuid behaviour for nano-sized helium droplets. Rota-
tionally-resolved infrared spectra of probe molecules such as
OCS in helium droplets conrm these ideas. In particular the
vibration–rotation spectrum of OCS in 4He droplets shows well-
resolved rotational structure similar to that in the gas phase
spectrum, whereas for 3He the rotational resolution for OCS is
completely lost. This marked difference between the two helium
11546 | Nanoscale, 2013, 5, 11545–11553
View Article Online
Feature Article
isotopes is evidence for superuid behaviour for 4He droplets
and normal liquid behaviour for 3He droplets.29
The very low temperature and the fact that the helium
remains liquid make these droplets unique reactors for the
growth of nanoparticles. Bulk liquid helium is unsuitable as a
medium for growing nanoparticles because the added
materials will rapidly condense on the walls of the liquid
helium container. On the other hand helium droplets have a
distinct wall-less boundary that connes dopants within the
droplets. Aer being picked up by helium droplets, which can
be achieved using a so-called pickup cell, the dopant(s) will be
cooled rapidly to the ambient droplet temperature (0.38 K)
with a cooling rate as high as 1010 K s
1.30 At such a low nal
temperature all substances condense and therefore materials
that are not normally used in fabricating nanoparticles, such
as many liquids and even gaseous materials at room temper-
ature, can now be added to helium droplets to form clusters
and nanoparticles. It is this almost unlimited exibility in the
choice of pickup material that makes helium droplets suitable
for forming a vast array of novel nanoparticles.
Other factors that impact on the synthesis of nanoparticles
in superuid helium include the mean size and the size
distribution of helium droplets along with the pickup condi-
tions employed for the dopant(s), i.e. the gas or vapour pressure
and the length of the pickup zone. Similar to clusters produced
in standard supersonic beam experiments, the size of helium
droplets formed by expansion of helium gas follows a statistical
distribution. Consequently, the precise size of individual
helium droplets cannot be controlled but the mean droplet size,
hNi (where N is the number of helium atoms), is controllable
through adjustment of the source conditions, i.e., the helium
stagnation pressure and temperature behind the nozzle. The
size of helium droplets sets the upper limit for the largest
nanoparticles that can be grown.
Because the pickup of dopants follows Poisson statistics,
even for a monodisperse source of helium droplets there
would be a range of nanoparticle sizes. Of course in practice
the distribution of nanoparticle sizes is compounded by the
spread of droplet sizes found in practical helium droplet
sources.19 Depending on whether the expansion conditions
are on the high-temperature or low-temperature side of the
isentrope in the helium phase diagram,18,31 helium droplets
are formed either from the aggregation of gaseous helium
(subcritical expansion) or by the break-up of liquid (super-
critical expansion), with the average size ranging from 103 to
1012 helium atoms.32 For subcritical conditions relatively
small helium droplets are formed (hNi # 3  104) where the
sizes follow a log-normal distribution. Much larger droplets
are formed in supercritical expansions and the size distribu-
tion here is of an exponential variety, where the probability of
a droplet with a particular size decreases exponentially
beyond a certain droplet size. An extreme case has been
observed when the helium behind the nozzle is cooled
below the superuid transition temperature. Under such
conditions ultra-large monodisperse helium droplets, with a
diameter roughly equal to that of the nozzle aperture, have
been formed.24
This journal is ª The Royal Society of Chemistry 2013

Feature Article
For the addition of dopants to helium droplets it is usually a
good approximation to equate the geometrical cross-section of a
helium droplet with the pickup cross-section, since experiments
show that the droplets have a near unity pickup probability for
collisions with most foreign species.33 When a dopant is added
to a helium droplet any internal energy and kinetic energy of the
former will be transferred to the latter and released through
evaporation of helium atoms, leading to a decrease in the initial
droplet size. Thus for a second pickup event the pickup prob-
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
ability will decline because of the reduced geometric cross
section of the droplet, although the effect will be small if the
loss of helium is minor relative to the overall helium content of
the droplet. The loss of helium atoms and the reduction of the
droplet size will tend to be more signicant when dopant
species interact strongly. For example, the addition of one Ag
atom to a pre-existing Ag cluster releases an energy of approxi-
mately 3 eV due to the formation of a Ag–Agn bond,34
resulting in the loss of 4800 helium atoms (assuming 5 cm
1 of
energy is taken away by each helium atom). Consequently, if the
initial helium droplet size is small it will take the addition of
only a few Ag atoms before complete evaporation of the helium
droplet occurs.
Another important issue to consider is where the dopant
resides, i.e. inside or on the surface of the helium droplet. So far
all molecules and many atoms have been found to locate
themselves in the interior of helium droplets aer being picked
up. Permanent residence at the surface of helium droplets only
occurs for some metal atoms with diffuse valence electron
distributions, such as the alkalis and some alkaline earth
metals.18,35 The preference for a surface location only arises
when the attractive part of the dopant-helium interatomic
potential is overwhelmed by the short-range Pauli repulsion.
Theory indicates that the alkali atoms locate themselves in a
dimple on the surface of a helium droplet which balances the
short-range repulsion and the long-range London force.36
Finally, it is noteworthy that a travelling helium droplet
beam can pick up more than one type of dopant when it passes
through a pickup region, either via addition of a premixed gas to
a single pickup cell or by the sequential addition of different
types of materials using a series of pickup cells. In the former
case one would obtain mixed clusters and/or particles, while for
the latter it is possible to form core–shell structures with the
upstream material as the core and the downstream material as
the shell. In this way helium droplets offer a straightforward
way of obtaining control of core–shell structures simply by
using two or more pickup cells and deciding on the order in
which the dopants are added.37–39
3. Formation of clusters and nanoparticles
in helium droplets
When more than one dopant atom/molecule is picked up by the
helium droplet agglomeration is expected. The agglomeration
process is driven mainly by long-range attractive forces between
the dopant species while the superuidity of the helium droplet
allows almost unhindered migration. However, another factor
to bear in mind is the low temperature of the helium and the
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
very rapid cooling, which can lead to the formation of unusual
clusters. A remarkable illustration of this was provided by Nauta
and Miller, who formed long-chain (HCN)n clusters (n ¼ 2–7) in
helium droplets and conrmed their structures by laser spec-
troscopy.40 Whereas HCN molecules tend to form more stable
structures such as the cyclic trimer in a conventional supersonic
expansion, in helium droplets higher energy structures, i.e., the
head-to-tail aligned linear clusters, are formed and are then
stabilized by the helium droplets. The explanation here is that
the HCN molecules are steered into position by long-range
dipole–dipole forces and become trapped in a shallow potential
energy minimum by the continuous rapid cooling, even though
a lower energy minimum is available.
A wide range of metal clusters have been synthesised in
helium droplets and have been investigated using mass spec-
trometry, including clusters of the alkali metals Na, K, Cs and
Rb,36,41–47 heteronuclear K–Rb clusters,48,49 the coinage metal Ag,
the group 13 metal In,50 and the transition metal Cr.51 Only in a
few cases have sizeable metal clusters composed of a few tens of
atoms in helium droplets been studied, such as Ag, Eu,52 Na,53
K,54 Cs and Rb.55 Alkali metal atoms can combine either with
anti-parallel spins, leading to the formation of strong metal–
metal bonds, or with parallel spins, in which case weakly bound
high-spin state van der Waals clusters result. For larger clusters
it has been suggested that the formation of spin-polarized
congurations is of low probability, so these clusters will most
likely be in their electronic ground states.56 One might therefore
expect that the chance of forming large alkali clusters on the
surface of helium droplets is remote, given the large energy
release when forming large low-spin alkali clusters. However,
when the metal cluster becomes sufficiently large other factors,
such as the interaction potential between the clusters and the
helium, come into play. Theoretical modelling has predicted
that the dispersion force increases faster than the repulsive
force and at some critical size the alkali cluster may become
submerged in the helium rather than adopt a surface location.47
For example, Nan clusters have been found to enter 4He droplets
once n $ 21, whereas Kn clusters require approximately
78 atoms to become fully solvated.53,54
Larger metallic clusters composed of over 1000 atoms have
also been produced in helium droplets, such as Cdn,35,57 Znn,35,57
and Mgn,57 and were observed via mass spectrometry. Although
these large metal clusters fall well within the nanoscale size
regime, with a diameter of several nm, only recently have
helium droplets been applied as reactors to specically
synthesise nanoparticles. This work has included Ag–Au nano-
particles composed of 500 metal atoms,58 pure Ag nano-
particles composed of over 6000 Ag atoms,59,60 Ag/Au and Ni/Au
core–shell nanoparticles,38 Mg nanoclusters,61 poly-crystalline
Ag nanoparticles,62 and Cu clusters.63
Crucial to most applications of nanoparticles formed in
helium droplets is the ability to so-land these particles on solid
substrates. The requirements for so-landing are illustrated by
the following example, which has been rehearsed by Loginov
et al.59 First, we note that the kinetic energy of the nanoparticle
as it approaches the substrate is determined by the mass of the
particle and the velocity of the helium droplets (typically 200–300
Nanoscale, 2013, 5, 11545–11553 | 11547

Nanoscale
m s
1, depending on the size of the helium droplets). Suppose
the helium droplets contain a silver cluster. We can then
calculate the kinetic energy per Ag atom as 0.034 eV if we assume
a droplet velocity of 250 m s
1. As pointed out by Loginov et al.,
this is far lower than the binding energy between a silver atom
and a silver cluster (3 eV), and it is also lower than the binding to
a TEM substrate (1 eV for amorphous carbon surface). The
deposition therefore falls well inside the regime of so landing
(dened as #0.1 eV kinetic energy per atom).64,65 In addition, the
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
evaporation of helium atoms when the droplets impact on the
surface will also contribute to a so landing. Aguirre et al. have
recently modeled the He–TiO2(110)-(1  1) interaction using
density functional calculations and applied this to the collision
of a helium droplet with a TiO2 surface.66 Their model predicts
that the helium droplet spreads on the surface as it collides and
leads to the formation of a thin lm above the substrate, indi-
cating that the helium layer can act as a cushion for the so-
landing of nanoparticles embedded within the droplet.
3.1 Typical apparatus for the synthesis of nanoparticles
A typical helium droplet apparatus for the synthesis of metallic
nanoparticles consists of a continuous helium droplet source,
oven evaporators, a nanoparticle deposition station, and diag-
nostic equipment such as a mass spectrometer. To illustrate
this we show a schematic diagram of a helium droplet system in
Fig. 1 that has recently been designed and constructed in
Leicester specically for nanoparticle formation. Briey, a
closed-cycle cryostat is used to pre-cool high-purity helium gas
(99.9999%) and helium droplets are then formed by supersonic
expansion of the helium into a vacuum through a pinhole
nozzle with an aperture of 5 mm diameter. The operating
Fig. 1 Schematic of the UHV helium droplet apparatus used for the synthesis of core–shell nanoparticles. (a) XYZ manipulator; (b) cryostat; (c) skimmer; (d) pickup
region, including 3 oven evaporators and 3 cryo-traps (CT); (e) deposition station composed of (e1) XYZ manipulator, (e2) sample holder, (e3) load-lock chamber and
(e4) magnetically coupled sample translator; (f) quadrupole mass spectrometer. GV ¼ gate valve.
11548 | Nanoscale, 2013, 5, 11545–11553
View Article Online
Feature Article
stagnation pressure can be varied from 10–50 bar and the
temperature can be accurately controlled from 3 to 30 K. The
cryostat is mounted on an XYZ manipulator, allowing ne
control of the position of the nozzle relative to the skimmer. In
addition gate valves are employed to isolate the helium droplet
source, the pickup region, the load-lock chamber and the mass
spectrometer when necessary, allowing independent mainte-
nance in each of these four sections without disturbing the rest.
Metal vapours are produced by evaporation from alumina
ovens, each with a length of 75 mm and an aperture of 6 mm
diameter centred on the axis of the helium droplet beam. These
ovens are heated resistively using tantalum wire wrapped
around the cylindrical alumina container (supplied by FRIATEC
AG), which can reach a temperature up to 1950  C, enabling the
evaporation of a wide range of materials. In addition, a cryo-trap
of length 60 mm and with an axial aperture of 10 mm is located
beyond each oven: each trap is cooled by liquid nitrogen and the
aim is to reduce the free diffusion of metal atoms into the
deposition region. The deposition rate can be measured by a
thin lm deposition monitor (not shown in the gure) and the
deposition targets can be accessed and manipulated through a
load-lock chamber.
There have been two general types of oven designs employed
in the synthesis of nanoparticles using helium droplets: either a
Knudsen cell in a crossed beam style of arrangement61 or an
axial pickup cell.38,59 The ovens are usually resistively heated but
in an interesting development an electron bombardment
system has also been developed.51,62 In this case the oven is
placed off axis so it is essentially a Knudsen cell. By electron
bombardment almost any solid materials can be evaporated,
and it can heat a sample even without any container if the
material is directly under bombardment. A Knudsen-style
This journal is ª The Royal Society of Chemistry 2013

Feature Article
effusion cell produces metal atoms translating across the
helium droplet beam, which can help to reduce the free diffu-
sion of metal atoms toward the deposition target, especially
when there is a cold target opposite the evaporator to collect
those metal atoms that are not picked up by helium droplets.
On the other hand, in such a design each atom may carry a
signicant momentum at high temperature, which can lead to
deviation of the helium droplets from their original axial
direction in the worst-case scenario. This deviation can be
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
avoided by using an axial pickup cell, as in our work, which
allows dopants to collide with helium droplets from any direc-
tion with equal probability, and hence when many dopant
atoms/molecules are added the net added momentum is zero.
However, with this approach diffusion of metal atoms along the
droplet beam axis cannot be entirely avoided. This means that
metal atoms from one oven can enter a subsequent oven and
might even eventually make their way through to the deposition
substrate.
To illustrate this we have placed a TEM substrate 20 cm away
from a gold evaporator, with the helium droplet source
temperature set at 20 K. At this temperature helium droplets are
formed but with an average size <1000 helium atoms, so are too
small to allow Au nanoparticles to form inside the droplets.
Aer a 5-minute deposition period nothing was seen on the
TEM substrate but aer 15 minutes very small nanoparticles
were observed on the surface of the substrate (see Fig. 2(a)). In a
second set of experiments the source temperature was set to 8 K,
at which point gold nanoparticles were produced given that
much larger helium droplets were present. Aer 5 minutes of
deposition gold nanoparticles with a diameter of 3 nm were
Fig. 2 TEM images of gold nanoparticles. (a) The temperature of helium droplet
source (T0) was set at 20 K and the deposition time (td) was 15 minutes; (b). T0 ¼
8 K and td ¼ 5 minutes; (c). T0 ¼ 8 K and td ¼ 15 minutes.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
clearly seen (see Fig. 2(b)) and only trace quantities of smaller
particles were observed. When nanoparticles were deposited for
15 minutes more than three times as many small particles were
observed than shown in Fig. 2(b). This clearly suggests that
small particles are formed by the aggregation of Au atoms on
the TEM substrate rather than inside helium droplets: other-
wise the quantity of small nanoparticles deposited aer
15 minutes should be exactly three times as many as those
deposited in 5 minutes. It is worth noting that in Fig. 2(c) the
average size of the small particles is clearly larger than those in
Fig. 2(a), which is due to the much higher ux of helium at T0 ¼
8 K and this can therefore carry metal atoms forward along the
droplet beam travel direction.
Nanoparticles formed in helium droplets can be investigated
in situ by a depletion spectroscopy technique,60,63 where the
optical absorption by the nanoparticle ultimately leads to
shrinkage of the helium droplet, due to heat dissipation. This
shrinkage leads to a drop in the electron impact cross section,
which can be used to register the absorption process in
combination with mass spectroscopy. In the case of coinage
metal nanoparticles such as silver and gold, the surface plas-
mon can lead to strong photon absorption, and this has been
shown to produce a substantial depletion effect allowing their
spectra to be recorded in helium droplets. To obtain the actual
size, shape and composition of the nanoparticles formed in
helium droplets, it is essential to use ex situ characterisation
methods such as XPS and TEM. One could build a helium
droplet instrument that incorporates dedicated XPS and TEM
facilities but this would be very expensive. It is far more cost-
effective to transport samples towards specialist analytical and
microscopy facilities in a vacuum-tight transportation station.
3.2 The major challenges, opportunities and perspectives
One of the major limitations of using helium droplets as a
means to synthesise nanoparticles is the deposition rate. The
typical ux of helium droplets that can be used to synthesise
nanoparticles is in the range of 109 to 1012 s
1, which sets the
upper limit for the ux of nanoparticles impinging on a
substrate surface. In a recent experiment Emery et al. have
measured the deposition rate of Mg clusters formed by the
addition of Mg atoms to helium droplets and have found a
deposition rate of 3–50 ng cm
2 s
1, depending on the size of
helium droplets.61 On these grounds it would be challenging to
produce much more than 100 mg of nanoparticles per day. For
fundamental research this may be sufficient in many cases, e.g.,
for the investigation of fundamental nanoscience, such as the
optical and magnetic properties of the particles. It may also be
sufficient for applications where a surface coating of the
material is required, e.g., for making sensors, for heterogeneous
catalysis, and for data storage based on arrays of nanoparticles.
However, if a much larger quantity of the nanoparticles is
required, as might be the case for some biological or medical
applications, there is a major challenge in scaling up the
process.
The surface location of some dopants, such as alkali metals,
also sets a challenge for the synthesis of nanoparticles
Nanoscale, 2013, 5, 11545–11553 | 11549
 View Article Online

Nanoscale Feature Article

containing these materials. As mentioned earlier the formation
of low-spin light alkali metal clusters will release a substantial
amount of energy (equal to that of the binding energy between
the two atoms in their ground electronic states), which can
result in facile ejection of the cluster from the droplet surface.41
To reduce the possibility of ejection, a possible solution is to
add these metals to helium droplets in the form of pre-formed
clusters rather than single atoms, which might be achieved by
use of a sputtering source. For heavier atoms such as Rb and Cs,
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
4  106 and 5  105 helium atoms, respectively. During these
experiments the stagnation pressure was 15 bar and the Au and
Ni ovens were maintained at a constant temperature. As shown
in Fig. 3(a), at 9.5 K only isolated and nearly spherical nano-
particles are present, with a mean diameter in the region of 2.5
nm. When the temperature is lowered spherical nanoparticles
of slightly larger diameter are seen, but now some rod-like
particles are also present. This is suggestive of the onset of
multi-centre growth, because the rod-like particles have the

it is uncertain whether the ejection of low-spin clusters is any same diameter as the spherical particles and seem to show an
longer a problem because of the relatively low metal–metal internal structure, which suggests that they are agglomerates of
binding energies.55 Calculations have shown a much lower 2–4 spherical particles. Going to even lower nozzle temperatures
binding energy (no more than 2000 cm
1) when an Rb or a Cs the rods become longer and we can identify quite long chains,
atom is added to an existing cluster and forms the lower spin with the longest chain observed being about 75 nm in length. It
state, which can be dissipated by the boil-off of #400 helium is not clear why the agglomeration leads to open, chain-like
atoms only. This means that Cs and Rb clusters are less likely to structures rather than more spherical entities. The formation of
‘boil off’ the droplet while growing, and indeed once sufficiently Ag chains in large helium droplets has recently been observed
large they may submerge into the droplet. and has been ascribed to cluster formation in quantized
Another challenge is associated with the size of helium vortices.67 However, at least in the case encountered in Fig. 3,
droplets. To grow the largest possible nanoparticles, one needs the chain formation seems more likely to result from multi-
the largest possible helium droplets. Very large helium droplets centre particle growth and subsequent particle agglomeration
can be formed by passing liquid helium, rather than gaseous into rods and chains.
helium, through the pinhole nozzle and into the vacuum The synthesis of nanoparticles using superuid helium is
system.24 When the temperature is sufficiently low the helium restricted to small quantities of nanoparticles until suitable
departs the nozzle as a microjet, which eventually ssures into scale-up techniques are developed. However, the amount of
droplets with a diameter on the order of the nozzle aperture particles that can be obtained currently is sufficient for funda-
diameter. These droplets ought to be called helium micro- mental research, as well as for applications that require small
droplets given their size and each may consist of somewhere in amounts of nanoparticles, e.g., the deposition of nanoparticle
the region of 1010 helium atoms. Inside these ‘super’ helium thin lms.61 As an illustration of this we have recently deposited
droplets it is possible in principle to grow nanoparticles with
diameters up to several hundred nm. However, the ux of
helium droplets will be now even lower than for the expansion
of gaseous helium and thus a lower production rate for nano-
particles will ensue.
An additional complication is that it is not straightforward to
obtain large nanoparticles using ultra-large helium droplets
because multi-centre growth will come into play. Indeed
evidence has been presented for multi-centre growth in droplets
composed of 107 helium atoms.60 This phenomenon was rst
reported by Loginov et al., who observed absorption spectra of
Ag nanoparticles in the infrared, mimicking the behaviour of
colloidal cluster–cluster aggregates of nanoparticles. This
indicates the growth of several small silver clusters in a single
large helium nanodroplet ([106 helium atoms) rather than a
single compact particle when a high Ag doping rate was used.
Similar behaviour was recently observed by the same research
group for Cun clusters with n > 1000.63 In such cases the
consecutive pickup events occur faster than the migration time
required for a metal atom to reach any existing particle near the
centre of the droplet. As a result, several small metal particles
can be formed at different nucleation sites within a large
helium droplet, which eventually agglomerate to given a nano-
particle with clear grain boundaries.
As a further illustration of multi-centre growth we have
Fig. 3 TEM images of Ni–Au nanoparticles formed by the sequential addition of
sequentially added Ni and Au to helium droplets with the nozzle Ni and Au to helium droplets at a source temperature of (a) 9.5 K, (b) 8 K, (c) 7 K
set at various source temperatures, with T0 ¼ 6, 7, 8 and 9.5 K, and (d) 6 K. The surface is populated by spherical nanoparticles at 9.5 K but
respectively, which gave the estimated droplet sizes of 108, 107, nanoparticle chains form at lower temperatures.

11550 | Nanoscale, 2013, 5, 11545–11553 This journal is ª The Royal Society of Chemistry 2013

Feature Article
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
Fig. 4 Image of a silver nanoparticle thin film deposited on a glass for 3 hours.
The nanoparticles were formed in helium droplets at a source temperature of 6 K
and stagnation pressure of 15 bar, giving an average helium droplet size of hNi
 108.
Ag nanoparticles for 3 hours on a glass substrate and formed a
thin lm (see Fig. 4). The average size of the nanoparticles was
found to be 6 nm diameter from TEM. According to the
measured deposition rate by Emery et al.,61 we could estimate
that over 20 layers of nanoparticles have been deposited
(assuming a deposition rate deposition rate of 10 ng cm
2
s
1), which is clearly visible under a white light source. Such
particle thin lms may nd niche applications in devices such
as sensors and solar cells.
Although there are limitations in using helium droplets as a
tool for synthesising nanoparticles, they also offer some very
exciting possibilities in both fundamental nanoscience and
nanotechnology. In the context of fundamental nanoscience a
wide variety of nanoparticles, including core–shell and core–
multiple shell nanoparticles, can be synthesised and can be
investigated via spectroscopy and microscopy outside of the
helium droplets. Furthermore, the synthesis of nanoparticles in
helium droplets also has a good degree of control of the size of
the core and the size of shell, so that nanoparticles with specic
functions can be fabricated, such as magnetic nanoparticles,
plasmonic nanoparticles and nano-catalysts. Once such parti-
cles have been characterised and their properties are shown to
be controllable one can begin to explore potential scale-up
techniques to increase the formation rate.
Finally, we propose two interesting experiments that can
signicantly broaden the utility of superuid helium in the
fabrication of nanoparticles. (1) Novel nanoparticles with a gas
or liquid as the core and a protective layer as the shell. With
superuid helium droplets any substance except helium will
condense; hence it is possible to incorporate substances that
are gases or liquids at room temperature into helium droplets
as the core, which can then be coated by a protective shell and
can be taken out of vacuum for investigations and applications.
This is an unprecedented approach that cannot be readily
implemented via any conventional nanoparticle synthetic
techniques. One can conceive of a number of applications of
this type of nanoparticle, such as the usage of these particles to
explore solution properties conned to the nanoscale. Equally,
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
one might try to trap one or more solute molecules within a
solvent in this nanoscale core. One could even imagine trapping
drug molecules in this environment and, at least conceivably,
such systems could be used for drug delivery.
(2) The formation of core–multiple shell nanoparticles. So far
there have been several examples of the successful synthesis of
core–shell nanoparticles but the formation of core–multiple
shell nanoparticles using superuid helium has not yet been
exploited. This should be straightforward to achieve using
helium droplets, i.e., by using more than two pickup cells. By
way of contrast core–multiple shell nanoparticles are very
difficult to make using conventional wet-chemistry techniques.
To give an example, we suggest a new category of magnetic
nanoparticles with a ferromagnetic core and an antiferromag-
netic shell, such as a Fe/Cr combination. Such nanoparticles are
expected to possess a very high coercive eld, unidirectional
anisotropy and giant magnetic moments. However, such parti-
cles are also highly susceptible to oxidation, which can be pre-
vented by coating the particle with an inert protective layer,
such as gold. A wide range of applications can be envisaged with
these nanoparticles, such as MR imaging, hyperthermia treat-
ment of cancer and magnetic data storage.
To close this section we summarize the key advantages of
helium droplets.
(1) Due to the very low temperature, any added substance will
attach to and condense within the liquid helium, providing it
can be delivered initially as a gas or vapour.
(2) Helium droplets can pick up dopants sequentially by
using a series of pickup cells and therefore offer the opportunity
to form core–shell and even core–multiple-shell nanoparticles.
(3) The low-temperature environment for the synthesis of
nanoparticles minimizes the inter-diffusion between layers, and
hence the expectation is that nanoparticles can be produced
with sharp boundaries between layers.
(4) Given that helium is inert it allows the growth of nano-
particles using highly reactive materials.
(5) A good degree of control on the size of nanoparticles and
the amounts of materials in the core and each layer can be
achieved by choosing the appropriate pickup conditions, i.e.,
the mean droplet sizes and the partial pressures of the dopants
in the respective pickup cells.
(6) Helium droplets offer the possibility of so-landing the
embedded nanoparticles onto a solid surface.
4. Conclusions
Helium droplets offer a unique environment for the formation
and investigation of clusters and nanoparticles. Here we have
attempted to give succinct descriptions of the properties of
these droplets, how they are formed and doped, and how their
properties combine to allow the formation of clusters. The
properties of helium droplets were then discussed in the
context of nanoparticle formation, which opens up new possi-
bilities for using helium droplets as a powerful tool in nano-
science and nanotechnology. By addressing this we hope to
encourage the scientic community to consider the many
exciting possibilities offered by nanoparticle growth in helium
Nanoscale, 2013, 5, 11545–11553 | 11551

Nanoscale
droplets and to recognise the advantages of this technique over
more conventional methods for nanoparticle formation.
Finally, we point out the major challenges to be faced when
synthesising novel nanoparticles using helium nanodroplets.
As these challenges are overcome we can expect to see appli-
cations in areas as diverse as catalysis, magnetic data storage
and sensor development, to name but three.
Acknowledgements
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
SY, CB and AME thank the EPSRC and the Leverhulme Trust for
grant support and the Advanced Microscopy Centre at the
University of Leicester for providing TEM facilities.
References
1 S. K. Murthy, Int. J. Nanomed., 2007, 2, 129.
2 X. Xia and Y. Xia, Front. Phys., 2013, DOI: 10.1007/s11467-
013-0318-8.
3 A. N. Shipway, E. Katz and I. Willner, ChemPhysChem, 2000,
1, 18.
4 T. J. Kempa, R. W. Day, S. Kim, H. Park and C. M. Lieber,
Energy Environ. Sci., 2013, 6, 719.
5 E. C. Corbos, P. R. Ellis, J. Cookson, V. Briois, T. I. Hyde,
G. Sanka and P. T. Bishop, Catal. Sci. Technol., 2013, DOI:
10.1039/c3cy00255a.
6 A. J. Gellman and N. Shukla, Nat. Mater., 2009, 8, 87.
7 N. A. Mashayekhi, Y. Y. Wu, M. C. Kung and H. H. Kung,
Chem. Commun., 2012, 48, 10096.
8 A. Popov, J. Lademann, A. Priezzhev and R. Myllyla, J. Biomed.
Opt., 2005, 10, 064037.
9 R. Kessler, Environ. Health Perspect., 2011, 119, A120.
10 R. J. Macfarlane, B. Lee, M. R. Jones, N. Harris, G. C. Schatz
and C. A. Mirkin, Science, 2011, 334, 204.
11 R. Ferrando, J. Jellinek and R. L. Johnston, Chem. Rev., 2008,
108, 845.
12 R. G. Chaudhuri and S. Paria, Chem. Rev., 2011, 112, 2373.
13 C. Granqvist, L. Kish and W. H. Marlow, Gas phase
nanoparticle synthesis, Kluwer Academic Publishers,
Dordrecht, 2004.
14 C. M. Wang, D. R. Baer, J. E. Amonette, M. H. Engelhard,
Y. Qiang and J. Antony, Nanotechnology, 2007, 18, 255603.
15 J. P. Toennies and A. F. Vilesov, Annu. Rev. Phys. Chem., 1998,
49, 1.
16 C. Callegari, W. Jäger and F. Stienkemeier, Handbook of
Nanophysics: Nanoparticles and Quantum Dots, CRC Press,
Taylor and Francis Group, 2011, P. 4-1.
17 K. Kuyanov-Pozument, D. Skvortsov, M. N. Slipchenko,
B. G. Sartakov and A. F. Vilesov, Physics and Chemistry at
low temperatures, Pan Stanford Publishing Pte. Ltd., 2011,
P. 203.
18 J. P. Toennies and A. F. Vilesov, Angew. Chem., Int. Ed., 2004,
43, 2622.
19 C. Callegari and W. E. Ernst, Handbook of High-Resolution
Spectroscopy, ed. M. Quack and F. Merkt, John Wiley &
Sons, Ltd., 2011, P. 1551.
20 S. Yang and A. M. Ellis, Chem. Soc. Rev., 2013, 42, 472.
11552 | Nanoscale, 2013, 5, 11545–11553
View Article Online
Feature Article
21 F. Pobell, Matter and Methods at Low Temperatures, Springer-
Verlag, Berlin, 1996.
22 J. Wilks and D. S. Betts, An Introduction to Liquid Helium,
Clarendon Press, Oxford, 1987.
23 D. R. Tilley and J. Tilley, Superuidity and Superconductivity,
Institute of Physics, Bristol, 1990.
24 R. E. Grisenti and J. P. Toennies, Phys. Rev. Lett., 2003, 90,
34501.
25 J. P. Toennies, A. F. Vilesov and K. B. Whaley, Phys. Today,
2001, 54, 31.
26 J. Harms, J. P. Toennies and E. L. Knuth, J. Chem. Phys., 1997,
106, 3348.
27 M. Hartmann, R. E. Miller, J. P. Toennies and A. Vilesov,
Phys. Rev. Lett., 1995, 75, 1566.
28 S. Grebenev, M. Hartmann, M. Havenith, B. Sartakov,
J. P. Toennies and A. F. Vilesov, J. Chem. Phys., 2000, 112,
4485.
29 S. Grebenev, J. P. Toennies and A. F. Vilesov, Science, 1998,
279, 2083.
30 G. D. Stein, Surf. Sci., 1995, 156, 44.
31 H. Buchenau, E. L. Knuth, J. Northby, J. P. Toennies and
C. Winkler, J. Chem. Phys., 1990, 92, 6875.
32 L. F. Gomez, E. Loginov, R. Sliter and A. F. Vilesov, J. Chem.
Phys., 2011, 135, 154201.
33 M. Lewerenz, B. Schilling and J. P. Toennies, J. Chem. Phys.,
1995, 102, 8191.
34 S. Krückeberg, G. Dietrich, K. Lützenkirchen,
L. Schweikhard, C. Walther and J. Ziegler, J. Chem. Phys.,
1999, 110, 7216.
35 J. Tiggesbäumker and F. Stienkemeier, Phys. Chem. Chem.
Phys., 2007, 9, 4748.
36 O. Bünermann, G. Droppelmann, A. Hernando, R. Mayo and
F. Stienkemeier, J. Phys. Chem. A, 2007, 111, 12684.
37 J. Liu, B. Shepperson, A. M. Ellis and S. Yang, Phys. Chem.
Chem. Phys., 2011, 13, 13920.
38 A. Boatwright, F. Cheng, D. Spence, E. Latimer, C. Binns,
A. M. Ellis and S. Yang, Faraday Discuss., 2013, 162, 113.
39 E. Loginov, L. F. Gomez and A. F. Vilesov, J. Phys. Chem. A,
2013, DOI: 10.1021/jp402614s.
40 K. Nauta and R. E. Miller, Science, 1999, 285, 1895.
41 J. Higgins, C. Callegari, J. Reho, F. Stienkemeier, W. E. Ernst,
K. K. Lehmann, M. Gutowski and G. Scoles, J. Phys. Chem. A,
1998, 102, 4952.
42 J. H. Reho, J. Higgins, M. Nooijen, K. K. Lehmann, G. Scoles
and M. Gutowski, J. Chem. Phys., 2001, 115, 10265.
43 A. W. Hauser, C. Callegari, P. Soldan and W. E. Ernst,
J. Chem. Phys., 2008, 129, 044307.
44 G. Auböck, J. Nagl, C. Callegari and W. E. Ernst, J. Chem.
Phys., 2008, 129, 114501.
45 A. W. Hauser, G. Auböck, C. Callegari and W. E. Ernst,
J. Chem. Phys., 2010, 132, 164310.
46 A. W. Hauser, C. Callegari, P. Soldan and W. E. Ernst, Chem.
Phys., 2010, 375, 73.
47 C. Stark and V. V. Kresin, Phys. Rev. B: Condens. Matter Mater.
Phys., 2010, 81, 085401.
48 J. Nagl, G. Auböck, A. W. Hauser, O. Allard, C. Callegari and
W. E. Ernst, Phys. Rev. Lett., 2008, 100, 063001.
This journal is ª The Royal Society of Chemistry 2013

Feature Article
49 J. Nagl, G. Auböck, A. W. Hauser, O. Allard, C. Callegari and
W. E. Ernst, J. Chem. Phys., 2008, 128, 154320.
50 A. Bartelt, J. D. Close, F. Federmann, N. Quaas and
J. P. Toennies, Phys. Rev. Lett., 1996, 77, 3525.
51 M. Ratschek, M. Koch and W. E. Ernst, J. Chem. Phys., 2012,
136, 104201.
52 J. D. Close, F. Federmann, K. Hoffmann and N. Quaas, J. Low
Temp. Phys., 1998, 111, 661.
53 L. An der Lan, P. Bartl, C. Leidlmair, H. Schöbel, R. Jochum,
Published on 02 October 2013. Downloaded by University of Leicester on 7/7/2021 11:06:47 AM.
S. Deni, T. D. Märk, A. M. Ellis and P. Scheier, J. Chem.
Phys., 2011, 135, 044309.
54 L. An der Lan, P. Bartl, C. Leidlmair, H. Schöbel,
R. Jochum, S. Deni, T. D. Märk, A. M. Ellis and
P. Scheier, Phys. Rev. B: Condens. Matter Mater. Phys.,
2012, 85, 115414.
55 M. Theisen, F. Lackner and W. E. Ernst, J. Phys. Chem. A,
2011, 115, 7005.
56 S. Vongehr and V. V. Kresin, J. Chem. Phys., 2003, 119,
11124.
57 T. Diederich, T. Döppner, T. Fennel, J. Tiggesbäumker and
K.-H. Meiwes-Broer, Phys. Rev. A: At., Mol., Opt. Phys., 2005,
72, 23203.
This journal is ª The Royal Society of Chemistry 2013
View Article Online
Nanoscale
58 V. Mozhayskiy, M. N. Slipchenko, V. K. Adamchuk and
A. F. Vilesov, J. Chem. Phys., 2007, 127, 094701.
59 E. Loginov, L. F. Gomez and A. F. Vilesov, J. Phys. Chem. A,
2011, 115, 7199.
60 E. Loginov, L. F. Gomez, N. Chiang, A. Halder, N. Guggemos,
V. V. Kresin and A. F. Vilesov, Phys. Rev. Lett., 2011, 106,
233401.
61 S. E. Emery, K. B. Rider, B. K. Little and C. M. Lindsay, J. Phys.
Chem. C, 2013, 117, 2358.
62 A. Volk, P. Thaler, M. Koch, E. Fisslthaler, W. Grogger and
W. E. Ernst, J. Chem. Phys., 2013, 138, 214312.
63 L. F. Gomez, E. Loginov, A. Halder, V. V. Kresin and
A. F. Vilesov, Int. J. Nanosci., 2013, 12, 1350014.
64 Metal Clusters at Surfaces, Structure, Quantum Properties,
Physical Chemistry, ed. K.-H. Meiwes-Broer, Springer,
Berlin, 2000.
65 Metallic nanoparticles, Handbook of Metal Physics, ed.
J. Blackman, Elsevier, Oxford, 2009.
66 N. F. Aguirre, D. Mateo, A. O. Mitrushchenkov, M. Pi and
M. P. de Lara-Castells, J. Chem. Phys., 2012, 136, 124703.
67 L. F. Gomez, E. Loginov and A. F. Vilesov, Phys. Rev. Lett.,
2012, 108, 155302.
Nanoscale, 2013, 5, 11545–11553 | 11553